LoadKaro is a hyper-local, real-time freight and cargo matching platform (Uber for Freight). Built for the HackIndia VibeCoding Hackathon 2026, it bridges the massive gap in India's unorganized local logistics sector by combining the price transparency of a Reverse Auction with the affordability of Cargo Pooling. It connects individual customers and B2B partners with truck drivers through an intuitive, dynamic, and fully responsive cross-platform application.
- Inefficient Consumer Logistics: Apps like Porter force users to pay for an entire Tata Ace (₹800+) even if they only need to move a single washing machine.
- Unorganized "Naka" Market: Haggling with local tempo drivers is time-consuming, lacks price transparency, and offers no safety guarantees.
- No Loading Assistance: Existing apps provide the vehicle but expect the customer to handle the heavy lifting.
- Bypasses expensive map APIs by utilizing Gemini 1.5 Flash AI as a blazing-fast geocoding engine.
- Features real-time location autocompletion and instantaneous coordinate-to-address reverse geocoding on pin drag.
- How it works: A customer posts a cargo requirement. Nearby verified tempo drivers bid down on the fare within a 3-minute window via real-time WebSockets.
- Impact: Eliminates negotiation stress and creates fair competition.
- How it works: Similar to ride-share, but for goods. Our ML algorithm groups overlapping origin/destination vectors.
- Impact: Two customers share one vehicle, each paying 40% less. Unlocks a new market of micro-cargo.
- Mobile (Android/iOS): Native map integration using OpenStreetMap (
<UrlTile>) requiring no Google Maps API Key. - Web: High-performance, interactive Leaflet map seamlessly integrated with React Native Web.
- Smart Routing Engine: Integrates with the public OSRM (Open Source Routing Machine) API to instantly calculate and draw real-world turn-by-turn road geometry. Calculates precise Indian fare estimates based on actual Haversine distance formulas.
- Add 1-2 verified helpers (gig workers/students) to the booking for ₹100-150/hour to provide end-to-end service.
- Specialized pre-mapped lanes and priority booking routes for registered enterprise partners. Cloud kitchens and retail stores can buy monthly delivery quotas.
- Framework: Expo / React Native (Universal codebase for Web, iOS, and Android)
- Navigation: Expo Router
- State Management: Zustand
- Maps:
react-native-maps(Mobile) & Leaflet (Web) - Styling: Modern, responsive flexbox with Uber-inspired minimalist design tokens
- Framework: FastAPI (Python 3.10+) for async, high-performance microservices.
- AI Integration: Google Gemini API (Geocoding & NLP Search), OpenCV (AR Load Validator).
- Routing & Spatial Data: OSRM HTTP API and PostgreSQL with PostGIS extension.
- Communication: REST & WebSockets (Redis Pub/Sub for the active auction state).
cd backend
python -m venv venv
source venv/bin/activate # On Windows use: venv\Scripts\activate
pip install -r requirements.txt
uvicorn main:app --reload --host 0.0.0.0 --port 8000
cd docker
cp .env.example .env
docker-compose up -dcd mobile
npm install
npx expo start
- Press
wto open in your web browser. - Scan the QR code with the Expo Go app to run on your physical Android or iOS device.
- Grant Location: Upon launch, the app grabs exact GPS coordinates and reverse-geocodes it via AI into a pickup street address.
- Select Drop-off: Drag the red pin on the interactive map or type an address in the smart search box.
- Get Route & View Prices: The app talks to OSRM to draw the shortest street route and calculates precise prices using the Haversine algorithm for both Solo and SahiYatri modes.
- Request: Confirm the ride to launch the load into the live WebSocket auction pool!
- Driver Bidding: Drivers in the zone see the load on their radar and compete by lowering their bids.
Captain: Arsh Chakraborty
- Contributors:
Arsh1233Chhaviii23
Built for the digital revolution of localized logistics.